At the Computex 2025 event, AMD introduced its new line of graphics cards designed for workstations, AI applications, content creators, and technical fields. The standout announcement was the Radeon AI PRO R9700, which features 32 GB of GDDR6 memory and the RDNA 4 architecture. This new card joins AMD’s Radeon AI PRO family.
The Radeon AI PRO R9700 is not aimed at gaming but rather at high-demand work environments, particularly in corporate and research settings. The card is built to handle local AI model running, model adjustments, and complex renderings. It doesn’t rely on cloud resources, which is important for privacy, low latency, and better control over data.
Key Features of the Radeon AI PRO R9700
- 64 Compute Units
- 32 GB GDDR6 VRAM
- Base Clock Speed: 2.35 GHz
- Boost Clock Speed: Up to 2.92 GHz
- Memory Interface: 256-bit
- Cache: 64 MB
- Power Consumption: 300 W
- Supports PCIe 5.0
The card also includes AMD’s second-generation AI accelerators, offering performance that is up to twice as fast as the previous generation. It supports multi-GPU systems, expanding memory and processing power for parallel tasks like simulations and simultaneous inferences.
The Radeon AI PRO R9700 is expected to be available from July 2025, although an official price has not been announced yet. The card will likely be aimed at high-performance workstations.
AMD’s Threadripper 9000 Series Also Gets a Boost
Alongside the new GPU, AMD introduced the Ryzen Threadripper 9000 processors, which include models like the 9995WX with 96 cores and 192 threads. These processors are designed for complex renderings, physical simulations, and training AI models. They come with up to 480 MB of cache and support for PCIe 5.0.
During the event, AMD highlighted the growing importance of embedded AI in PCs, especially in professional and business environments. AMD is partnering with companies like ASUS and Lenovo to bring AI features to new workstations and notebooks. The goal is to integrate up to 50 TOPS of AI performance in these devices.
As AI continues to advance, the demand for optimized hardware solutions in gaming, creation, and model development is increasing rapidly.
